´óÏó´«Ã½

and have just blogged about a ´óÏó´«Ã½ project they've just started working on that will in Matt's words will open up a

a huge (´óÏó´«Ã½) database that’s been carefully tended by a gang of crack ´óÏó´«Ã½ librarians for decades. Nearly a million programmes are catalogued, with descriptions, contributor details and annotations drawn from a wonderfully detailed controlled vocabulary.

This should be in public beta early in the New Year, but if you have any ideas or comments based on the screenshots and explanations from Ben and Matt then let them/us know.

Ben Hammersley:
Matt Biddulph, HackDiary:
UPDATE:
Tom Loosemore has also blogged his thoughts on this:

• Bookmark with:
• |
• |
• - What's this?

Tom Coates has posted some mock ups, notes and screen shots of the ´óÏó´«Ã½ Radio and Music teams .
It contains some fascinating new ways for users to contribute, add context and leave their traces on audio files and radio programmes and has already been much

Tom explains the context behind it..

An on-demand archive is going to make the number of choices available to a given individual at any point almost completely unmanageable. And then there's the user-generated content - the amateur and semi-professional creations, podcasts and the like that are proliferating across the internet. In the longer term there are potentially billions of these media creators in the world. All of this choice, however, creates some significant problems - how on earth are people expected to navigate all of this content?"

The ideas expressed in the prototype (and the way its been implemented by several of Tom's team) is fascinating stuff. We'd be really keen to see any feedback or leave comments on Tom's blog on your thoughts on this.

This was Tom's last project for the ´óÏó´«Ã½. We'll really miss him. He was a frequent visitor, supporter and critic! (in a good way) of Backstage.

By the way his (ex) music colleagues have also let us know that they have started to make the 1000+ reviews on the ´óÏó´«Ã½ Music site available for syndication.
This is the RSS feed.

• Bookmark with:
• |
• |
• - What's this?

A while back now we asked you to .

The response was magnificent and thanks again to everyone who took the time and effort to submit entries. But on to the bit you really want:

The winners of the first Backstage competition are...

Leon Brocard and Leo Lapworth for

We all loved this.

It’s a fantastically impressive prototype with a wealth of features and innovative ideas already implemented that we were almost overwhelmed. That said the simplicity of the user experience meant that it was easy to use out of the box. There is a lot of different findability here; tagging, ratings, recommendations as well as the basic grid format with innovative and different ways of navigating your way through thousands of hours of TV and radio.

There was a real effort to integrate non ´óÏó´«Ã½ feeds (imdb, bleb) and the TV Anytime metadata was neatly exploited via the accessibility search tool as well as genre lists such as films and comedy. I haven't mentioned the personalisation (but you can sign up for your preferred channels) and some real heavy lifting behind the scenes plumbing like an open source PERL TV-anytime parser which the team have made available to CPAN. We were itching to use this and show it around as soon as it was submitted which is a sure sign that Leo and Leon had come up with something special...

This is fantastic prototype which thoroughly deserves to win the first Backstage competition.
Leon and Leo win some "Geek Bling", ie: a rackmount server to the cost of £1K and a trip to the ´óÏó´«Ã½ to talk further.

Runners Up:

Thomas Scott for the TV Map

It was the lateral approach to the brief that we liked here. The idea that TV metadata descriptions could be versioned as a google map mash up sounds odd but, in fact, works very well.

This is a very simple prototype with only a weeks sample data but we can see the opportunities for introducing viewers to programmes they would have never considered viewing.

It’s rough and ready but does the job of suggesting that there are different ways to visualise those hundreds of tv and radio programmes per week. Thanks Thomas.

Fraser Nevett for the Programme Similarity Visualiser

The complete departure from the grid was what made this stand out. There's a simple algorithm there combined with a compact visualisation tool and Fraser has done a fine job of representing a weeks broadcasts in a vastly different way to the time/channel axis. There were plenty of ways in which we could think how this could be built on with plenty of different features but it was again another easy to use prototype with an extremely intuitive interface. More importantly it set us thinking about what might be possible.

We'll be dropping a note to the winners/runners up to arrange prizes/delivery and meet ups as soon as this mail has arrived

Please don't think that because your prototype wasn't shortlisted that we weren't impressed. There was some great use of phones/sms, avatars, automated speech, some traditional but keeping it simple listings layouts, some fine personalised epgs and a Radio 4 clock!. Thanks for putting the effort in.

Watch out for further competitions...

• Bookmark with:
• |
• |
• - What's this?

Fast search tool for ´óÏó´«Ã½ RSS news feeds.

• Bookmark with:
• |
• |
• - What's this?

Buliding on my previous Greasemonkey script/web site ()
I have now put the data into Google Earth Format.

This is my first attempt at a Google Earth file so it isn't perfect. Please email me any suggestions/comments/problems (see site for email address).

It includes data from the ´óÏó´«Ã½ Travel Feeds for roads and rail/rube delays. It also includes links to Live Departure Boards for railway stations.

• Bookmark with:
• |
• |
• - What's this?

The ´óÏó´«Ã½ regularly advertise free tickets for audiences to attend recordings of various TV and radio shows. These are currently advertised at /whatson/tickets/ - a page which I'm always forgetting to visit.

What's more, once word of a popular show gets around, tickets get snapped up quickly. Apparently some Little Britain tickets recently went in as little as 4 hours.

There are also seperate pages for the 3 ´óÏó´«Ã½ Nations (Scotland, Wales, and Northern Ireland) listing other recordings. These are linked to from the main page in the What's On section.

The idea? A pretty simple one. RSS feeds of the above, listing each recording as a seperate item. The feed would then contain a link to more information about the show, and details of how to apply for tickets.

• Bookmark with:
• |
• |
• - What's this?

Charlotte is a natural language interface to the TVAnytime scheduling information. You can ask her in plain english questions like "when is spooks on", "what is on tomorrow night" and so on. For added impact we've given her an avatar and text-to-speech system - but these aren't essential to her function.

Further work will include more patterns for her AIML matching system, better handling of tag searches ("eg what gritty dramas are on this week"), more natural presentation of results, and conversely, a list mode when you just want a blow by blow schedule.

Charlotte currently focusses on ´óÏó´«Ã½1 and 2 only.

• Bookmark with:
• |
• |
• - What's this?

I listen to Radio Four a lot. For me - and I suspect many others who share my addiction - the schedule has a dual purpose. It tells me what's on, but also how far through the day we are. This clock is for everyone who leaves the house by 'Thought for the Day' and feeds the cat at quarter-past 'Front Row'.

You need an SVG viewer (such as ). Some viewers may have difficulty with the javascript - there's a tick box at the bottom to turn it off. You can also adjust the time zone and choose a different station to view.

It's built in Java with JSP. A daily cron task on my server downloads and unzips the feed. The JSP page that produces the clock refreshes its cache of programme details each day. I use the open source TVAnytime parser to handle the XML, creating simple 'programme' objects with title, date, summary and url fields to manipulate. Positioning the titles on the SVG needed just a little trigonometry, as did the initial angle of the three hands. The rotation uses the animateTransform tag with varying durations.

With more time, I'd like to
- avoid the javascript issue
- look at colour coding programmes by genre
- write the summaries in interesting ways (round the clock face?)
- offer a range of configurable designs
- offer direct 'listen again' links where possible

• Bookmark with:
• |
• |
• - What's this?

A prototype to take the ´óÏó´«Ã½ backstage 7-day schedule data, convert it to RDF and use it to build a dynamic and personalised web view. Our aim is to meet the ´óÏó´«Ã½ aspiration for a "cool, original and innovative [utility]"

WHY? To learn something about the use of RDF for open data storage and query; to build an AJAX application; to provide a useful and fun service.

HOW? the schedule data is provided as XML, so an XSLT transform turns this into RDF, which is served using Joseki. The client application uses the AJAX approach, using a standard Javascript library.

WHAT NEXT: Performance! appears to be largely javascript (client) rather than Joseki (server). Collect user feedback. We've a list of wish-list features about a mile long to prioritise. Oh, and it'd be polite to get it working on Microsoft IE :)

CREDITS: Chief Hackers were Paolo Castagna, Richard Cyganiak and Damian Steer. Ideas and encouragement from Steve Cayzer, Mark Butler and others in HP Labs Bristol.

• Bookmark with:
• |
• |
• - What's this?

This program displays a grid showing "what's on" on the freeview channel lineup (TV & Radio). The difference is that it also shows how much it expects you to like the show, on a scale from -9 (my idea of hell is being forced to watch 5 minutes of this show) to +9 (I would kill to be able to see 5 minutes of this show). It does this by allowing you to rate shows, and using a vector-search algorithm to see how similar each show is to the totality of shows that you have rated.

If I had more time, I'd optimize the site more, tune the weightings of the vector search, possibly add phrases instead of treating every word of the description sepperately, and perhaps most importatly provide documentation and missing polish.

• Bookmark with:
• |
• |
• - What's this?

This prototype is, to put it simply, a TV listings system. It is based on a PHP backend, but uses AJAX to display all the data onto a grid similar to the one used by programs such as Digiguide.

The actual grid is fully scrollable, and it only gets the relevant information when needed (scroll to the end and it gets the next results). This is obviously limited by the data that is in the database.

If you click on a television or radio programme, you get the description of it, information about the genre and when it is repeated, as well as some extra information that may be relevant to a few people (aspect ratio, etc).

If I had the time (it has been a busy few months), there would be many other features in it, such as the ability to search by genre and search for a specific show, and add your own favourites into it so that you know what is on on a specific day that you want to watch or listen to.

I do plan to actually use this myself, so most of those features will end up in it sometime.

(More about this entry can be found at )

• Bookmark with:
• |
• |
• - What's this?

Using the ´óÏó´«Ã½ schedule data this program allows the schedule to be searched and links each program by their group and genre. All 19 TV and Radio channels are included.

The data is displayed in a simple list format with the short synopsis displayed when the mouse is moved over the item, the more detailed information from the feed is displayed when an item is selected, example subtitles, audio description, widescreen, closedcaption & sign language. this also includes alternate schedule information if the program is repeated within the 7 day window.

Future ideas to be developed would be to allow users to add programs, groups and genre's to their favourites and to be sent reminders when certain programs are due to be broadcast.

• Bookmark with:
• del.icio.us |
• Digg |
• Reddit
• - What's this?

Presenting the TVMap: mixing up the ´óÏó´«Ã½'s TV listings (and Bleb's feeds for other channels) with Google Maps, to produce a map that shows what locations will be on TV soon. Interested in all programs about Abu Dhabi? No problem, it'll email you about them. Want to know if there's anything about Canada on today? Sure, not a problem.

This first prototype version is showing a static display for October 3rd, but in the next day or so it'll get features like showing up to six days ahead, and email and RSS location-based alerts ("let me know when there's a program within 50 miles of Vienna").

It works by parsing out all proper names from the descriptions of programs and running them against a subset of world GNIS data -- it won't pick up every reference, and right now a couple of things have to be cleaned by hand (for example, Chucklevision doesn't take place *in* Barry...), but some quick filters should sort that out in the next couple of days.

Whew. Talk about cutting it fine. 40 minutes until the competition deadline, but I've just managed to finish the first prototype! Tomorrow - the good stuff gets added!

Cheers,

Tom

• Bookmark with:
• |
• |
• - What's this?

mightyv () lets you find and rate your television programs, rearrange your schedule, find recommendations, fetch RSS feeds for your series or schedule and more.

Find out more at

• Bookmark with:
• |
• |
• - What's this?

My idea started with looking at making TV Listings accessible. This is why I came up with the combination of using Web, SMS and Phone services to provide the information.

I've created a simple web site that allows you to browse the TV and Radio listings for the next week. This allows you to create your own personal listings that you can receive alerts on via either email or SMS. This web site is written in PHP accessing a MySQL database. The database is populated using the Perl TV-Anytime module once a day.

If you phone 01483 604620 you will be provided with a menu system that will allow you to navigate the programs on any of the ´óÏó´«Ã½ TV Channels. This information is updated every 10 minutes by process that discovers which programs are running by accessing the database and this text is then used to generate voice using Cepstral software. When required the Asterisk phone system plays these sound files back to anyone phoning that number.

The SMS service accepts messages on 07914 229975 and will then respond with the information you require. Eg. Sms 'TV ´óÏó´«Ã½1' and after a few minutes you will get a response with the next few programs on that channel. This is using gnokii over Bluetooth to connect to the phone and Perl script are handling the processing and sending of the text messages.

There are numerous improvements. As this is only a prototype a more reliable service for sending SMS would be required and the VOIP services is only running on a ADSL connection, so the sound quality can be poor. See for more information and ideas.

• Bookmark with:
• |
• |
• - What's this?

From the guy who brought you "Rebotcast reads ´óÏó´«Ã½ News | World | UK Edition", I now bring you "Rebotcast Reads ´óÏó´«Ã½ Tv Anytime Listings". This is a speech-synthesized reading of the listings. For my prototype, I have limited my scope to the "Documentary" genre in the radio listings.

This code owes a sincere debt to Leon Brocard, who wrote a series of Perl modules to parse TV Anytime listings.

The podcast is updated every day by downloading the last listings, using a "cron" job. There are some obvious improvements to make, such as including the times of each radio program. I just wanted to get a placeholder up there before the deadline.

Another improvement, more long term, would be to give users the ability to design their own feeds, by inputting parameters in a web form.

• Bookmark with:
• |
• |
• - What's this?

Installing the Prototype:

This prototype has been written in Java so you’ll need the java Runtime environment to run the application. You can get this at

To install this prototype just download the .rar file and unpack it to any folder. After unpacking go to the ´óÏó´«Ã½ folder and run "start epg.bata".

In this show case the EPG information included is from 26-9-2005 to 4-10-2005. If you want to see the TVguide after this date, you should add the new files which you can find at
Just download the latest file, unpack and copy them into the ´óÏó´«Ã½\xml directory

The Prototype:

Because of the large amount of available channels these days, there has to be a selection to adjust the number of your favourite TV and radio channels shown in the EPG on your TV. It gives the viewer the option to make a selection of his favourite ´óÏó´«Ã½ channels which can be altered at any time. In the environment of ´óÏó´«Ã½i it is possible to create your own Personalised ´óÏó´«Ã½i- EPG.

This application is a prototype of a personalised EPG; you can choose whether you want to see the entire guide, or just a guide which shows the channels you selected. This prototype demonstrates the concept working into the ´óÏó´«Ã½i environment; the final idea is to get the application working on OpenTV-and MHP set-top-boxes.

Further there is an option to activate alerts. These alerts will be covering a selection of genres. While it's not fully functional yet, this option gives a good idea of how it should work. Just like the personal channel selection the viewer can select the genres he is most interested in. When viewing the personal or full EPG the EPG will highlight the programmes in those genres. the viewer will be able to see at a glance which programmes are within his interests.

The application uses the TV-Anytime API made available by the ´óÏó´«Ã½ and the Listings in the TV-Anytime format also made available by the ´óÏó´«Ã½.

Future Ideas:
Since the alert function is not fully functional at this moment this should off course be finished. At this moment the personalisation is at channel level. But with some more advanced filtering it should be able to filter at programme level, or even segment level. The meta data of TV anytime gives a lot of opportunities for alerts in your own Personalised ´óÏó´«Ã½i EPG.

The next step will be to create your private profile. On basis of your profile you will get TV programme suggestions.

The EPG data is now downloaded manually, but for a future version it should be possible to download them over the Internet automatically. In the end the entire application should be ported to MHP and OpenTV.

• Bookmark with:
• |
• |
• - What's this?

The Unofficial ´óÏó´«Ã½ 7 Day TV and Radio Listing is a online radio and TV listings that uses the ´óÏó´«Ã½ TV and Radio feeds provided by backstage.bbc.co.uk. A Java program parses the backstage XML data and inserts it into a mySQL database each day and a PHP based website displays all of the information.

The website displays full 7 day listing for each channel. I f you choose a specific program it display all of the information for that program as well as links to the ´óÏó´«Ã½'s information for that program. If it is a movie links to IMDB and Google also appear. Movies also have a simple 5 star ranking system associated with them so that users can rate the movie.

You can add favourite programs so that they are highlighted on the listings.

You can search by keyword, genre, or actor name to find any programmes that match your criteria. The search can also be narrowed down by selecting a specific date or date range and a channel.

There is a basic forum for programmes that allows users to talk to each other about programmes.

When a user starts adding favourites recmmendations will start being generated for them based upon what other people that have the same favourites as you also watch.

Therre are a few things on the site that need to berefined still. The CSS needs a bit of work to be fully compatiable with IE and also for smaller screen resolutions. The advanced search also has a few minor bugs to be ironed out.

In the future I would like to output statistics about the kind of programs that different people like (eg. Favourite programs for men and woman, and also for age renge etc..). I would also like to write a desktop client that uses the favourites data and alerts a user when their favourite program is about to start. I would enhance the forum section to include ´óÏó´«Ã½ode and a word censor to hide swear words. I have also toyed with the idea of letting users tag programs and then adding a search option for tags.

• Bookmark with:
• |
• |
• - What's this?

There are so many dictionaries out there on the net - all giving you the meaning of a word.

My idea is to use the ´óÏó´«Ã½ data to show people how the word they search is used in conversation or writing. You can even group this by years to see how the meaning of the word has changed over time - context and grammer.

This will be a more useful service for people using dictionaries - e.g. I like the word, but how and where can I use it correctly.

I hope i finish my prototype in time ... I doubt it - just finished the dictionary ;)

• Bookmark with:
• |
• |
• - What's this?

If there were an easy way to find out CRIDs, such as a bookmarklet that generated a popup search screen, then this would enable them to be freely spread around in the same way that URIs are.

This is the scenario. I want to blog about last night's Spooks. I click on the button in my browser toolbar and up pops a dialog with fields for programme title and a rough timeframe ("in the last week"). It returns a URI which is copied to the clipboard.

I can then paste this in my blog post. I can make it a link to del.icio.us or Technorati so that I can tag the programme. Blog search tools could find who else is talking about the same series or programme.

It would also facilitate the linking of programme/series websites to the episodes. The page for last night's Spooks could embed it's CRID so that it takes ownership of that programme.

If I had the time I'd do a bit of Perl to do the searching. Once you've got the CRID then the rest is up to the public.

• Bookmark with:
• |
• |
• - What's this?

Thank you to everyone who has entered the . I’m pleased to announce that Matt Locke - Creative Director, ´óÏó´«Ã½ New Media & Technology – has kindly agreed to help judge the entries. Both myself (Ben Metcalfe) and Jem Stone from the backstage.bbc.co.uk team will also be assisting Matt with the judging.

There’s been some great work produced, which will definitely make our jobs even harder!

We will be announcing the winner sometime next week, so watch this space!

• Bookmark with:
• |
• |
• - What's this?